Percolation problems, which aim to find the critical volume fraction Vc of conductive particles in an insulating matrix, are mainly solved by analytical-mathematical models. So far, the resolution of the percolation threshold only considers primarily geometrical parameters as well as the conductivity of the particles and therefore assumes a homogeneous distribution of the particle. This consequently does not take into account the effect of the nature of the polymeric matrix on the arrangement of the particles which has been observed experimentally. This project, focused on the elaboration of a highly electrically conductive coating for aerospace applications, aims at obtaining with molecular simulations the conductivity of a material depending on the nature of the polymer and the type of the metallic particles (size, shape, nature). 

To be able to achieve that, a leading study on the interface metal-polymer is established. This investigation required the calculation of interfacial tensions of solid-liquid, solid-vapour and liquid-vapour interphases. From these characterisations, it was highlighted that the solid-liquid interactions must be re-adjusted to reproduce the work of adhesion. By tweaking the cross-terms of the Lennard Jones potentials, the experimental contact angle was successfully reproduced by molecular simulations. 

Finally, the transition to mesoscopic simulations was initiated. The use of the Statistical Trajectory Matching (STM) method was considered relevant to establish the coarse-grained models allowing the reproduction of a reference atomistic trajectory in a mesoscopic simulation. Subsequently, coarse-grained simulations with a larger set of particles are proposed to determine a percolation threshold influenced by the nature of the polymer. This will hopefully provide inputs to the differences in electrical behaviour of the formulated materials.